[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAD2nsn1a9q3n3dED1L6bomgsf1fP5TA9Tb9v+q9j=xjqdO2i3Q@mail.gmail.com>
Date: Wed, 18 Jan 2012 11:02:24 +0530
From: Mohammed Shafi <shafi.wireless@...il.com>
To: MR <g7af0ec1e3ea1e7b1@...tmail.ru>
Cc: linux-kernel@...r.kernel.org, linux-wireless@...r.kernel.org,
Rajkumar Manoharan <rmanohar@....qualcomm.com>
Subject: Re: ath9k crash 3.2-rc7
2012/1/18 MR <g7af0ec1e3ea1e7b1@...tmail.ru>:
> > > Turned out I got lucky previous time.
> > >
> > >This time I went to the other AP, spent some time under it, roamed
> back
> > and
> > >
> > >WiFi didn't disappear. I didn't even lose XMPP connections in the
> > process.
> > >
> > >
> > >Maybe debug printing (which prints a lot) managed to introduce some
> > extra
> > > delays into some race condition.
> > >
> > > I hope I will have time to test this more on Monday. Tomorrow I will
> be
> >
> > > using WiFi less than usual.
> >
> >Now I went a few time back and forth, spent a lot of time under
> previously
> >
> > risky AP and still there are no problems..
> >
> > I wonder if maybe a congested AP could help.. But unclear how to obtain
> > that quickly and reliably.
>
> Even with debugging off WiFi works reliably across different conditions.
>
you can test frequent roaming with the attached script. also you can
test roaming, after enabling ASPM(attached script, courtesy: Luis)
please completely read and understand
http://linuxwireless.org/en/users/Documentation/ASPM
incorrect tweak may cause hardware issues.
--
shafi
Download attachment "roam-script.sh" of type "application/x-sh" (2008 bytes)
Download attachment "enable-aspm" of type "application/octet-stream" (7310 bytes)
Powered by blists - more mailing lists